7,021 research outputs found
Curriculum learning for heterogeneous star network embedding via deep reinforcement learning
No full textLearning node representations for networks has attracted much attention recently due to its effectiveness in a variety of applications. This paper focuses on learning node representations for heterogeneous star networks, which have a center node type linked with multiple attribute node types through different types of edges. In heterogeneous star networks, we observe that the training order of different types of edges affects the learning performance significantly. Therefore we study learning curricula for node representation learning in heterogeneous star networks, i.e., learning an optimal sequence of edges of different types for the node representation learning process. We formulate the problem as a Markov decision process, with the action as selecting a specific type of edges for learning or terminating the training process, and the state as the sequence of edge types selected so far. The reward is calculated as the performance on external tasks with node representations as features, and the goal is to take a series of actions to maximize the cumulative rewards. We propose an approach based on deep reinforcement learning for this problem. Our approach leverages LSTM models to encode states and further estimate the expected cumulative reward of each state-action pair, which essentially measures the long-term performance of different actions at each state. Experimental results on real-world heterogeneous star networks demonstrate the effectiveness and efficiency of our approach over competitive baseline approaches.Submission published under a 24 month embargo labeled 'U of I Access', the embargo will last until 2020-08-01The student, Meng Qu, accepted the attached license on 2018-04-28 at 03:27.The student, Meng Qu, submitted this Thesis for approval on 2018-04-28 at 03:32.This Thesis was approved for publication on 2018-05-01 at 14:12.DSpace SAF Submission Ingestion Package generated from Vireo submission #12548 on 2018-09-27 at 11:15:31Made available in DSpace on 2018-09-27T16:28:00Z (GMT). No. of bitstreams: 2
QU-THESIS-2018.pdf: 796856 bytes, checksum: 30a59b6e036a099beb5aab939947cd7d (MD5)
LICENSE.txt: 4204 bytes, checksum: 719b2eb6b187103bda6cfc701f3c93be (MD5)
Previous issue date: 2018-05-01Embargo set by: Seth Robbins for item 107731
Lift date: 2020-09-27T16:28:07Z
Reason: Author requested U of Illinois access only (OA after 2yrs) in Vireo ETD systemEmbargo set by: Seth Robbins for item 107731
Lift date: 2020-09-27T16:30:34Z
Reason: Author requested U of Illinois access only (OA after 2yrs) in Vireo ETD systemEmbargo set by: Seth Robbins for item 107731
Lift date: 2020-09-27T16:31:43Z
Reason: Author requested U of Illinois access only (OA after 2yrs) in Vireo ETD systemEmbargo set by: Seth Robbins for item 107731
Lift date: 2020-09-27T16:34:29Z
Reason: Author requested U of Illinois access only (OA after 2yrs) in Vireo ETD systemU of I Only Restriction Lifted for Item 107731 on 2020-09-28T09:15:24Z
Rates of Return to University Education: The Regression Discontinuity Design
No full textEstimating the rate of return to a university degree has always been difficult due to the problem of omitted variable biases. Benefiting from a special feature of the University Admission system in China, which has clear cutoffs for university entry, combined with a unique data set with information on individual National College Entrance Examination (NCEE) scores, we estimate the Local Average Treatment Effects (LATE) of university education based on a Regression Discontinuity design. To the best of our knowledge, this is the first study to use RD design to estimate the causal effect of a university education on earnings. Our results show that the rates of return to 4-year university education relative to 3-year college education are 40 and 60 per cent for the compliers in the male and female samples, respectively, which are much larger than the simple OLS estimations revealed in previous literature. Since in our sample a large proportion of individuals are compliers (45 per cent for males and 48 per cent for females), the LATEs estimated in this paper have a relatively general implication. In addition, we find that the LATEs are likely to be larger than ATEs, suggesting that the inference drawn from average treatment effects might understate the true effects of the university expansion program introduced in China in 1999 and thereafter.rate of return to education, regression discontinuity design, China
Rates of Return to University Education: the Regression Discontinuity Design
No full textEstimating the rate of return to a university degree has always been difficult due to the problem of omitted variable biases. Benefiting from a special feature of the University Admission system in China, which has clear cutoffs for university entry, combined with a unique data set with information on individual National College Entrance Examination (NCEE) scores, we estimate the Local Average Treatment Effects (LATE) of university education based on a Regression Discontinuity design. To the best of our knowledge, this is the first study to use RD design to estimate the causal effect of a university education on earnings. Our results show that the rates of return to 4-year university education relative to 3-year college education are 40 and 60 per cent for the compliers in the male and female samples, respectively, which are much larger than the simple OLS estimations revealed in previous literature. Since in our sample a large proportion of individuals are compliers (45 per cent for males and 48 per cent for females), the LATEs estimated in this paper have a relatively general implication. In addition, we find that the LATEs are likely to be larger than ATEs, suggesting that the inference drawn from average treatment effects might understate the true effects of the university expansion program introduced in China in 1999 and thereafter.Rate of return to education, Regression Discontinuity Design, China
Pi pa gu qu
No full textSide A : 古琴music by 二梁 ; 琵琶. 1. 夢蝶 / 梁越銘作曲. 2. 壯士行 / 梁在平作曲. 3. 塞上曲. 4. 春江花月夜 -- Side B. 1. Pian interview.Live recording.Possibly reproduced from other commercial recording or radio broadcast (Pending for review)"台北故宮博物院"--Spine."77.元.9"--Index.Electronic reproduction from Rulan Chao Pian Audio Cassette Collection.Various composers.Performers, unknown.Spoken in Chinese."Taibei gu gong bo wu yuan"--Spine."77. yuan. 9"--Index.Side A : Gu qin music by er Liang ; pi pa. 1. Meng die / Liang Yueming zuo qu. 2. Zhuang shi xing / Liang Zaiping zuo qu. 3. Sai shang qu. 4. Chun jiang hua yue ye -- Side B. 1. Pian interview
Sources of variability in metro train-induced vibration
No full textIn previous in-situ measurements of metro trains it has been found that the velocity level on the track or tunnel wall may vary significantly between different train passages, even though the measuring section, the type of trains and the track and tunnel conditions are identical. An investigation is carried out into the sources of this variability, using a 3D train-track numerical model. This is built using the software SIMPACK and ABAQUS, and is connected through one-way coupling to a finite element model of the tunnel and soil. These models are used to study the influence of various train parameters, including the wheel and rail unevenness, train speed and degree of train loading. For comparison, in-situ measurements were made of the dynamic response of the rail and tunnel wall. The rail roughness at the site as well as the wheel unevenness of all 48 wheels for one train were measured. The results from the model indicate that the wheel unevenness affects the rail velocity level in the frequency region between 25 and 250 Hz and tunnel wall vibration above 5 Hz. The rail velocity level can vary by up to 20 dB due to wheel unevenness, with the largest variations occurring in the frequency bands 50–63 Hz. Variations in passenger loading affect the train-induced vibration by up to 4.5 dB, mainly in the low frequency region. When the train speed varies within a range of ±20% relative to the nominal speed 60 km/h, the frequencies of the peaks are shifted and the level in some frequency bands can change by as much as 10 dB. However, the largest influence is that of the wheel unevenness. It is concluded that the variation in these parameters, especially the wheel and rail unevenness, should be considered to achieve reliable predictions of train-induced vibration
A Profile of Fatherhood Among Young Men: Moving Away from Their Birth Family and Closer to Their Child.
No full textHave things changed all that much in terms of how fatherhood is conceptualized and exercised in daily life? That is the question underlying this article. The author compares the findings of a recent analysis on certain aspects of the lived experiences of young fathers (under 25 years of age) with the results of studies undertaken over the past ten years, and replies in the affirmative. First of all, when considering the representations held of fathers or mothers, most of these young fathers believe that their role is a multi-faceted one, and that it is often identical to that of their spouse. According to young fathers, fatherhood is a dual experience that requires them to be present on a daily basis while also casting their eye on the future. This is an experience that is constructed out of affectionate moments, child-care duties, education in the literal sense, and especially out of shared experiences with their spouse. In addition, they question the degree to which involvement in a career should take precedence over involvement in their child's life. In other words, the former 'competes' with their ability to be present in their child's daily life, which denotes a change from the attitudes of previous generations.Fatherhood, Young Fathers, Representation, Paternal Identity, Qualitative Research
取譬訓蒙 Qu pi xun meng.
No full textContient : I 教要總說 Jiao yao zong shuo ; II Qu pi xun mengNumérisation effectuée à partir d'un document de substitution.Catéchisme en exemples. Par le P. Zottoli. Publié à Zikawei (1870) avec l'autorisation (1869) du P. Agnello della Corte, Jésuite (1819-1896 ; noms Chinois Gu Zhen sheng Xian luo). Préface sans date
sj-pdf-1-imr-10.1177_03000605211035107 - Supplemental material for Snare-assisted precutting and dual-knife fistulotomy performed during difficult biliary cannulation in a patient with an ectopic papilla of Vater
No full textSupplemental material, sj-pdf-1-imr-10.1177_03000605211035107 for Snare-assisted precutting and dual-knife fistulotomy performed during difficult biliary cannulation in a patient with an ectopic papilla of Vater by Hui Liu, Chun-Meng Jiang, Bo Qu and Zhi-Guo Wang in Journal of International Medical Research</p
- …
